Co-Founder of Facetopo
Global Citizen Big Science Data Project
It's a big scientific question. All of our faces are blending: races are blending and we are seeing migration patterns. Now that we have genomics, which is accessible and people can get their genome sequence, a big question is can your face be a proxy for getting your genome sequence?
By Heidi Legg
In a large brick building filled with lofts on the corner of Harrison and Waltham Street in Boston’s South End, artists, tech startups, film studios, and the creative class have taken over. Along the third floor hallway of one of the SOWA art and design district offices, handmade signs and general artistry with arrows line the exposed brick walls, until we arrive at a large barn door that slides open into the light-filled studio where Alberta Chu bustles around a long wooden table.
Chu is well known in the Boston contemporary art scene as a supporter of all things where tech and art collide. She and her husband, scientist Murray Robinson, who has works in Boston’s genome discovery space, have teamed up to create a big data citizen science project using a mobile app to gather the data they need to explore the possibilities of facial morphology. Murray’s brother was diagnosed a decade ago with a rare disease, called Smith-Magenis Syndrome, that presents with facial markers, much like Down Syndrome. Chu and Robinson wonder, through the collection of millions of 3-D faces, if they will discover that the face is another gateway to the prediction of genetic discoveries. Does ethnography relate to your facial features?
I sat down with Chu to discuss what she is trying to build, the face as a gateway, and how society adapts to new technology.
What is Facetopo?
Facetopo is a global citizen science face-mapping project, and a big data project about faces, where we use an app as the way to gather the data.
Every user who wants to participate creates a private account and is able to download the app on either IOS or Android, where we provide instructions on how to move your face so that you can create a 3-D face map. Only you can access your face map when you log in.
Given we are a big data project about faces, we're building our own database for research. We have almost 4,000 faces from everywhere, but we need millions of people in order for this to be an actual big data project. We've only used organic marketing and word of mouth, so we're hoping with the fun selfie-feature people will share it and get their friends to do it. Somehow, it's spreading. People are curious.
How many faces do you need to even begin to move forward with your findings?
Our milestone goal is 10,000 faces. Once we have that, we will publish our first face tree in the form of a dendogram, which is a way of looking at large amounts of biological information. It's a tree-shaped model that shows evolution. Our dendogram will mathematically compare a similarity among different faces.
What exactly are you measuring or tallying?
There's no color or race information. Basically, we create a topography map of your face using about 80 dots [on the face], and then we measure the distance between each and every one of the dots. The dots are in the same place on everyone's face, but there are different measurements in between them. Using these measurements, we'll compare the mathematical differences and similarities among faces.
Is this a massive pure research scientific project or does it have a commercial application?
It's scientific but it also has a social aspect. It's a mobile app and maybe it will be something that's going to connect people who look like them or share measurements. If they want to be connected, they can opt in and maybe trade pictures, or eventually, find a twin. Your face is genetic so there could be similarities between you and someone who looks very much like you… but because it is science research, we don't yet know!
What we do know today is that genes are connected to behavior and disease. We don't yet know the accuracy of our mapping, but there may be groupings and that is why we are starting to do this as a citizen science project. We need to gather hundreds of times more data to see what will come out of it.
Why not just use the zillions of selfies posted online?
We are doing 3D face mapping; selfies are 2-D. When you try our app, you will see that we actually have to have the person moving their face. This is similar to what Apple is doing with their iPhone 8, but theirs is for security.
Won’t that mean Apple will get millions of faces pretty quickly, with their new face-mapping feature to unlock their iPhone 8? Does their announcement change things for you?
We're using our data for quantifying faces and users opt in to join our citizen research project. Apple is using face mapping like they used your fingerprint to unlock your phone. One is research and the other security. When you use Facetopo, there's a privacy agreement and a user policy, and you're the only one that can access your account. It’s yours and it's private. All you’re doing is donating the mass and the dot formation of your face to our big database, which is then analyzed. Nobody knows it's you. It's not connected to your identity, whereas Apple knows it’s your phone. It's got your name on it. The phone has a number on it.
Who else is working on Facetopo?
Facetopo is a project I cofounded with my husband Murray Robinson, who is in genomics with a career in cancer research. He's the genome person and scientist. The idea came out of a couples’ talk at Microsoft’s Nerd Night years ago at the Middlesex Lounge in Kendall Square. Nerd Night is a series of lectures where people present on their area of expertise.
Murray was set to speak about genomics but he wanted to make it fun and applicable, so we teamed up. He spoke about the genomics of faces, something of great interest to him because he has an older brother that has a genomic syndrome. Back in the '70s, there was no official name for it other than mentally handicapped, and it hadn't been classified when he was young. Ten years ago two scientists, Smith and Magenis, figured out his brother Kelly had something only 5,000 people have, now called Smith-Magenis Syndrome. These people have certain face markers, a fascination with electronics, and gregarious personalities. And it is genetic; he has a deletion of Chromosome 17.
So you and Murray believe the face is a gateway to medical research?
Yes, it could be a gateway to genomics or a gateway to learning something about yourself, which is why people are interested in their genome. It's all there, that recipe for who you are, and we could discover a great deal. We can see height and eye color and that kind of thing that is genetic when we look at a person, but fascination with electronics as a subgroup? That's a little surprising. But for people with Smith-Magenis, this is a genetic behavior.
What was your focus in the talk at Microsoft Nerd Night?
I talked about the connections between art and science through the ages, and different art movements that emerged from those intersections. We started working on this Facetopo project soon thereafter.
I began by collaborating with an engineer and designing a way to create a 3D-face scan using a mobile phone. I noticed that at the orthodontist office, they have those expensive cameras where your head is stationary and clamped in there, and the camera moves around you. We wondered if we could design a user experience with the mobile phone. We really wanted it to be on a phone because everyone has one today.
We've been in the iTunes store since March. We have created a fun selfie-filter when you make the Facetopo. We are encouraging users to take the 3-D image three times to ensure they're more accurate.
Given the new iPhone X is using face-mapping as a way to unlock your phone, there is new debate about this as a security feature and what else could be done with your face data. How is what you are doing different?
Ours is a global citizen project and a big science data project. With Facetopo, we think it's really important that given it’s your face – it's private and you control it. Your face isn't connected with your identity at all. It's really only a bunch of dots. The measurements could be anyone. It's basically eighty white dots that are in a 3D formation of a face. When you donate your face to the big data project, it's all numbers and dots and we can’t trace it back to who's who. It doesn't have your name on it.
What's the end goal?
It's a big scientific question. All of our faces are blending: races are blending and we are seeing migration patterns. Now that we have genomics, which is accessible and people can get their genome sequence, a big question is can your face be a proxy for getting your genome sequence?
When we have findings, we will publish a scientific paper.
Is facial morphology a cutting edge place right now for science?
Craig Venter published about genomics of faces in PNAS.org. [Venter and his colleagues at his company Human Longevity, Inc. did publish his research claiming to predict people’s physical traits from their DNA on Sept. 5, 2017 with The Proceedings in the National Academy of Sciences. Since then, his research based on 1,061 people of varying ages and ethnic backgrounds has been highly criticized as overblown by others in the field.] He already has a big database with his genome project and I think he's looked at 2,000 faces. He says that by looking at the genome, you can tell what someone's face is going to look like. They're still deciphering facial morphology and which genes do what, but people are starting to be interested in this. It’s been used for the diagnosis of Down Syndrome, Smith-Magenis Syndrome, and a handful of genetic syndromes where the face is a highly visible marker. This means that in people who don't have genetic syndromes, there could also be markers in your face that we don't know yet. That's why we're collecting data about faces. We don't know what we're going to find but you have to gather the data first to look at it.
Given your interest in art and science, is this pure science research for you or is there something more?
It is science but I think it's the kind of science that can connect people to others – perhaps even in the form of social media.
Murray and I would talk about getting our genomes done and I didn't want to do it. Given he is in genetics, he did it right away. He was one of the first 1,000 people to be sequenced. Given that the face is genetic, could it be a proxy? That’s the big scientific question, but we can't answer that question unless we collect a lot of data.
With faces there's this little creepiness factor you have to overcome, but I think people will become more comfortable.
This reminds me of the interview I did in 2014 with Harvard’s Professor of Genetics, Dr. George Church. He wants us all to give our genomes to his human genome project for similar research initiatives.
He loves Facetopo. It's exactly what they're doing with his Personal Genome Project collecting DNA, but with faces.
The public still seems resistant to give over its genomes. Why are we all so scared?
I think with any new emerging technology, it gets adapted by society. Maybe there's resistance at first but then it just becomes a part of us and it changes us. As these technologies become more and more invisible, such as when people are using their 3-D face to unlock their iPhone, everyone's going to be doing this all the time.
You believe society responds and adapts?
Yes, I think society adapts to new technologies. At Facetopo, we're protecting privacy and giving the power to the user to control where their faces are or aren't published, but I suppose the creepy part of facial recognition and surveillance is when there's some database of you. Today, if you’re getting off the tube in London, there are cameras on every corner and they use it to catch criminals. But then it's also a basic right to have privacy, and I get that you can't be anonymous anymore. We don't have that kind of surveillance society here in the US.
Our Facetopo data is not going to be shared with any outside databases. It's private. If people want to participate in scientific research that could lead to advances in behavior and disease, that’s pretty cool. There could be something in your face that could show something about you that’s really interesting, but in order to figure that out we have to have a lot of faces to compare. We're trying to make it interesting and fun for people.
Your face is out there; you've given it to Facebook and you don't control it. That is how we have differentiated in our how we designed our experience. With us, it's private and you log in and it's yours. And then in the back end, if your face matches with some other face, you'll have the choice, once we’ve developed more features, to either opt in or not. I mean, it's your face!
Are you looking for funding?
We will probably be looking for funding but I think we need to have some findings first.
Right now it’s a scientific study, but it could have fun social implications. You're comfortable with your family and your tribe and people who look like you, right? If you think back to early tribal days, people across the lake who were not like you and did not look like you could be your enemies, if they are not part of your tribe. You may have more in common with someone who looks similar to you.
What do you do with the data?
We will analyze the base data mathematically, create a dendogram and group things together by similarity, then compare and combine faces that are similar to make face groupings and morph faces to represent groups, which scientists are doing already. People who study psychology and faces have studied trustworthiness and perceptions of leadership, and we probably could do this with our facial measurements.
Will you ask for more data about our personalities and values?
It would be by questionnaire. The genome company, 23 and Me, does a lot of this.
Are you mining that information from people now?
No. We're not doing any surveys, but it may be something we'll do in the future.
Why do you need to be fourteen to participate?
Your face changes at puberty. There's the child face and the adult face, and fourteen is generally the age where that happens.
Where do you go to relax?
Oh, I never relax. I need to take a nap right now. I put on a big art show that opened on Friday called The Nasty Woman Art Show, and it was a huge amount of work. We had seven hundred people come through to see work from almost 300 artists. I'm a little brain dead. All the funds from the sales went to Planned Parenthood and Color of Change. It was a resistance art exhibit to raise money for causes that need it in the face of the current administration.
Where do I go to relax? I bike. I bike all the way to the Esplanade from the South End with my kids.
Where is the edge in Boston right now for new ideas?
I think there's a lot of exciting stuff happening with public art in Boston right now. The city doesn't do a lot of funding of it, but I think there are producers of public art that are able to raise money and there's a critical mass building in Boston around the arts and innovation and how that is a cycle with technology. I think it’s something that Boston has on other places, and it could be the thing that Boston does better than anyone else.